an integrative framework of cooperative advertising
Post on 28-Jul-2022
6 Views
Preview:
TRANSCRIPT
An Integrative Framework of Cooperative Advertising: Should
Manufacturers Continuously Support Retailer Advertising?I
Guiomar Martın-Herran1,∗, Simon P. Sigue2,∗∗
Abstract
A two-period game is developed in a bilateral monopoly where, besides pricing decisions,
the retailer and manufacturer can set their advertising and cooperative advertising support
rates for each period. It is demonstrated that, in addition to the established continuous coop-
erative advertising programs, in which the retailer advertises and the manufacturer supports
retailer advertising in each period, two other advertising schedules are possible. First, the
retailer advertises in each period, while the manufacturer only supports the second-period
advertising. Second, whether or not the manufacturer provides a cooperative advertising
program in the first period, the retailer only advertises in the second period and receives
advertising support. The conditions under which each of these advertising arrangements is
implemented are identified. In a continuous cooperative advertising schedule, the manufac-
turer may change his advertising support over time depending on the nature of the long-term
effects of retailer advertising. The implications of these findings are discussed. Keywords:
Cooperative advertising. Game theory. Pricing. Retailer advertising.
IWe thank two anonymous reviewers for their helpful comments and Janice Thiessen for copy-editingassistance. The first author’s research is partially supported by MEC under projects ECO2011-24352 andECO2014-52343-P, co-financed by FEDER funds and the COST Action IS1104 “The EU in the new economiccomplex geography: models, tools and policy evaluation”.
∗Departamento de Economıa Aplicada (Matematicas), Universidad de Valladolid, Avda. Valle de Esgueva,6, 47011, Valladolid, Spain. Tel.: +34 983 423330 Fax: +34 983 423299; E-mail: guiomar@eco.uva.es
∗∗Corresponding author: Faculty of Business, Athabasca University, 201-13220 St. Albert Trail, Edmonton,AB, T5L 4W1, Canada. Tel.: 780 460 0340, E-mail: simons@athabascau.ca
1IMUVA, Universidad de Valladolid2Faculty of Business, Athabasca University (Canada)
Preprint submitted to JBR December 20, 2017
1. Introduction
The optimal design of cooperative advertising programs over time remains a major chal-
lenge for both scholars and decision makers. Cooperative advertising is a joint promotional
arrangement, whereby a manufacturer reimburses a percentage of advertising expenditures
that retailers support in promoting his product. A cooperative advertising program aims
at providing additional incentives to retailers to increase their local advertising of a manu-
facturer’s product. Retailer advertising is believed to benefit manufacturers in three ways.
First, retailers have a better knowledge of their local markets and can therefore undertake
more effective advertising programs for manufacturers’ products. Second, retailers use local
media, which generally apply lower advertising rates than do national media. Finally, retailer
local advertising is known to stimulate immediate sales at the retail level, although its long-
term effects on sales remain controversial (Jørgensen et al., 2000, 2001, 2003; Herrington and
Dempsey, 2005).
Two research streams have investigated the optimal design of cooperative advertising pro-
grams to retailers (See Aust and Buscher (2014) and Jørgensen and Zaccour (2014) for review).
The first research stream uses static games. The optimal strategies derived from these static
games apply to a single period and overlook, among others, the now well-established long-
term effects of retailer advertising (e.g., Berger, 1972; Huang and Li, 2001; Karray, 2013,
2015; Karray and Amin, 2015; Karray and Zaccour, 2006; Li et al., 2002; Szmerekovsky and
Zhang, 2009; Xie and Ai, 2006; Yan, 2010; Yan and Pei, 2015; Yan et. al., 2016). The findings
of this research stream are known to be more relevant in circumstances where the decision
environment is relatively stable and channel members’ decisions do not have carryover ef-
fects (Jørgensen and Zaccour, 2014). As a consequence, it is implicitly believed that channel
members’ decisions related to cooperative advertising do not change over time.
The second research stream uses sophisticated dynamic models and seeks to study more
realistic cooperative advertising situations where the environment can change and channel de-
cisions can have long-term effects. Most works in this second research stream uses differential
games (e.g., Jørgensen et al., 2000, 2001, 2003; He et al. 2011; Sigue and Chintagunta, 2009;
Zhang et al., 2013, 2015). For mathematical tractability, however, these works have stud-
ied infinite horizon cooperative advertising contracts and mainly derived stationary feedback
strategies, as they all use time-independent parameters. Not surprising, these works mainly
prescribe constant cooperative advertising support rates that do not change over time.
In their recent review of cooperative advertising works, Jørgensen and Zaccour (2014)
pointed out the exclusive prescription of stationary cooperative advertising rates as a serious
2
shortcoming of the current literature. In the real world, cooperative advertising programs, as
many other promotional activities, are offered within limited time periods and their support
rates are barely constant over time. Among other strategies, manufacturers change their
cooperative advertising contributions depending on seasonal periods and the type of local
advertising they want to stimulate. A manufacturer may choose to support retailer advertising
exclusively when the sales in the industry are at the seasonal peak. Similar practices are known
in the advertising literature as pulsing, when advertisers alternate between zero and positive
advertising levels (Sasieni, 1989; Villas-Boas, 1993). For example, Mitsubishi Motors (2012)
developed a three-month cooperative advertising program in 2012 that went from April 3 to
July 2. Honda Canada Inc. (2010) has a flexible annual cooperative advertising policy, which
allows special rates to support specific dealers’ sales initiatives. On the other hand, while
cooperative advertising programs set very specific requirements for the use of funds, they are
generally flexible and give enough freedom to retailers to use or not to use them during a
given period. As a matter of fact, while all authorized Mitsubishi Motors North America
dealers were eligible to participate in the 2012 cooperative advertising program, only those
who endeavored to meet the program requirements were able to take advantage of it.
There is therefore a need to further explore what drives the changes in both coopera-
tive advertising programs and retailer advertising schedules over time. This paper hopes
to contribute an integrative framework that can explain observed practices in the business
world and provide useful guidelines to help implement more effective cooperative advertising
programs. On the theoretical ground, this paper helps to integrate some of the findings of
previous static and dynamic cooperative advertising models in bilateral monopoly contexts.
Following these works, we develop a stylized two-period model in which a manufacturer sells a
single brand to a retailer. In each period, the manufacturer determines the optimal wholesale
price and cooperative advertising support rate, while the retailer sets the optimal rate of local
advertising and retail price. This setup allows various cooperative advertising and retailer
advertising schedules to be considered as potential equilibria. Technically, unlike current dif-
ferential games-based models that prescribe a constant cooperative advertising rate over time,
in our proposed configuration, the manufacturer may or may not offer cooperative advertising
support from one period to another. In response, the retailer may or may not advertise in a
given period even if cooperative advertising support is provided. The research questions then
are:
1. Should the retailer continuously advertise? And should the manufacturer continuously
support retailer advertising?
3
2. What are the drivers of change in the players’ strategies over time?
3. What types of cooperative advertising arrangements should the manufacturer and re-
tailer implement over time if they both act so as to maximize their individual profits?
To address these questions, we use the Stackelberg solution concept to derive equilibria
in a game where the manufacturer is assigned the leadership role. This research differs from
previous works in several ways. First, we disregard competition, which is studied in several
recent works and proven to affect cooperative advertising decisions, to focus on vertical
interactions over time (Karray, 2015; Karray and Amin, 2015; Yan and Pei, 2015; Yan et al.,
2016).
Second, we consider unconstrained cooperative advertising programs for which the man-
ufacturer does not set a maximum contribution to his cooperative advertising program as a
percentage of the retailer’s purchases. The 2012 Mitsubishi Motors cooperative advertising
program referred to above is a good example of a constrained cooperative advertising program.
It has been recently demonstrated that, even in a bilateral monopoly context, when such a
constraint is used, an increase of the manufacturer’s cooperative advertising support may not
translate to an increase of retailer advertising as is otherwise expected (Zhang et al., 2015).
In this particular case, our work is more in line with the majority of previous cooperative
advertising works.
Third, unlike previous static cooperative advertising models, this paper acknowledges the
possibility of retailer advertising carryover, which means that the first-period retailer advertis-
ing may also impact on the second-period demand. Therefore, unless the long-term effects of
retailer advertising are set to zero, the game played in the second period cannot be considered
as a mere successive static game.
Finally, compared to infinite horizon cooperative advertising contracts previously studied
in the literature, which exclusively lead to constant cooperative advertising rates, this paper
shows that cooperative advertising rates can change over time to support various types of
retailer advertising. As a matter of fact, we do not make any restrictive assumption on
the role of retailer advertising and its effects, as in some of the published works. Instead,
we study the general scenario where, depending on the content of this type of advertising,
retailer advertising can have no, negative, and positive carryover effects. Previous works do
not simultaneously investigate these three possible effects (He et al., 2014; Jørgensen et al.,
2000, 2001, 2003; Sigue and Chintagunta, 2009). In this paper we show that these effects play
a critical role in how channel members schedule their advertising decisions.
The remainder of the paper is organized as follows. Section 2 describes the model and
4
discusses its assumptions. Section 3 describes the methodology and derives the game equilib-
rium solutions. Section 4 studies how the manufacturer’s and retailer’s decisions change over
time. Section 5 compares the findings derived in Section 3. Finally, Section 6 concludes and
discusses the managerial and theoretical implications of our findings.
2. The model
Consider a bilateral monopoly in which a manufacturer enters into an exclusive distribution
arrangement with a retailer who then sells the manufacturer’s product to consumers. The
manufacturer’s product faces no direct competition or, when competition does exist as in the
automobile industry, it is disregarded to focus on how vertical interactions between channel
members affect advertising decisions. Also, because of the exclusive distribution arrangement,
there is no intrabrand competition at the retail level. As a matter of fact, the existence of
such vertical interactions in advertising has lately led companies such as Toyota and Honda
to prohibit certain types of retailer advertising that are believed to damage their brand image
(Cole, 2015). Let ai and si be, respectively, the rate of retailer’s local advertising and the
manufacturer’s cooperative advertising rate or the percentage of the retailer’s advertising
expenditures that the manufacturer is committed to share in period i, i ∈ {1, 2}. Also, let
wi and pi denote the wholesale and retail prices in period i, i ∈ {1, 2}. We consider that the
retail price is the effective price consumers pay for the product in period i.
As in many other papers in the distribution channel literature (Chu and Desai, 1995;
Martın-Herran et al., 2010; Sigue, 2008), we assume the following linear demand functions:
q1 = g − p1 + αa1 and q2 = g − p2 + βa1 + αa2. These demand functions are mainly used
for convenience and tractability. They can also approximate quite well more complicated
functions for both non-durable and durable products (Lilien et al. 1992). The parameter
g is positive and represents the baseline demand at the start of the game. For simplicity,
consumer sensitivity to retail prices in the two periods is normalized to 1. The parameters α
and β respectively represent the short-term and long-term effects of retailer advertising. The
short-term effects of retailer advertising (α) in the two periods are identical and positive. On
the other hand, the long-term effects of the first-period retailer advertising (β) on the second-
period demand can be either zero, negative, or positive depending on the type of retailer
advertising and the target market. For instance, β can take negative values when retailer
advertising hurts the brand image or goodwill (Jørgensen et al., 2003). Conversely, β can
takes positive values when the first-period retailer advertising contributes to enhancing the
brand image or goodwill (Jørgensen et al., 2000). β is set to zero when retailer advertising
5
does not have any impact on the second-period demand. This type of retailer advertising is
known as promotional advertising, as it exclusively stimulates short-term sales (Jørgensen et
al. 2000). Consistent with the empirical findings of the work by Herrington and Dempsey
(2005) in the automobile industry, we do not make any a priori assumption on the relative
importance of the short-term and long-term effects of retailer advertising. Either effect could
be higher than the other depending on the nature of retailer advertising. The baseline demand
of the second period is given by: g+βa1 and depends on the retailer’s first-period advertising.
Table 1: Model specification
Period 1 Period 2
Manufacturer’s controls w1, s1 w2, s2
Retailer’s controls p1, a1 p2, a2
Demand functions q1 = g − p1 + αa1 q2 = g − p2 + βa1 + αa2
Manufacturer’s profits M1 = w1q1 − 12s1a
21 M2 = w2q2 − 1
2s2a
22
Retailer’s profits R1 = (p1 − w1)q1 − 12(1− s1)a21 R2 = (p2 − w2)q2 − 1
2(1− s2)a22
The current model specification (see Table 1) assumes that the manufacturer does not
undertake any national advertising and fully relies on the retailer’s advertising. This assump-
tion is somehow simplistic as, in many cases, manufacturers do not relinquish all advertising
responsibilities to retailers (Sigue and Chintagunta, 2009). However, it allows us to focus
on retailer advertising and its short-term and long-term effects on the design of cooperative
advertising programs over time, which are the focal points of this study.
We assume the following convex costs for retailer advertising, 12a2i . Convex costs are
generally used in the literature when advertising decision variables enter linearly in the de-
mand functions (e.g., Chintagunta and Jain, 1992; Jørgensen et al., 2001; Martın-Herran and
Sigue, 2011). They mean that the marginal costs of advertising are increasing. Given the
manufacturer’s cooperative advertising arrangements, the manufacturer’s portions of retailer
advertising costs are, 12sia
2i , while the retailer’s effective advertising costs are 1
2(1− si)a2i .
Without loss of generality, we assume the production and other administrative costs to be
zero. The retailer’s (Ri) and manufacturer’s (Mi) profits in period i, i ∈ {1, 2}, are given by:
R1 = (p1 − w1)q1 −1
2(1− s1)a21, R2 = (p2 − w2)q2 −
1
2(1− s2)a22,
M1 = w1q1 −1
2s1a
21, M2 = w2q2 −
1
2s2a
22.
6
The manufacturer and retailer aim to maximize their respective profits over the two pe-
riods: M = M1 + tMM2 and R = R1 + tRR2, where tM and tR are the manufacturer’s and
retailer’s discount factors. The discount factors of the two players are normalized to 1 (tM =
tR = 1). The use of a common discount rate for channel members is standard in the literature
(e.g., Jørgensen et al., 2001, 2003; Jørgensen and Zaccour, 2014) and captures the idea that, in
a bilateral monopoly in which a retailer only sells a manufacturer’s product, the two channel
members have the same interest in future profits. It is also reasonable to assume that, in the
current context of very low interest rates, channel members do not heavily discount future
profits.
3. Equilibria
The Stackelberg equilibrium concept is used to derive the model equilibria. The manufac-
turer and retailer are, respectively, the Stackelberg leader and follower. The allocation of roles
is done on an ad hoc basis, but the manufacturer’s leadership is known to benefit all channel
members (Jørgensen et al. 2001). The sequence of moves of the two players is as follows.
Table 2: Sequence of moves
Player Decision variables
Stage 1 Manufacturer w1, s1
Stage 2 Retailer p1, a1
Stage 3 Manufacturer w2, s2
Stage 4 Retailer p2, a2
The manufacturer announces his first-period decisions in Stage 1. The retailer reacts to
the manufacturer’s announcement in Stage 2 to determine her optimal first-period strategies.
Subsequently, the manufacturer announces his second-period decisions in Stage 3, which is
followed by the reaction of the retailer in Stage 4 as she determines her optimal second-period
retail price and advertising rate. In such a configuration, subgame-perfect equilibria are
obtained by deriving optimal solutions backwards. Essentially, the retailer’s optimal second-
period strategies (p2, a2) in Stage 4 are first obtained. Second, the manufacturer’s optimal
second-period decisions (w2, s2) in Stage 3 are derived after the introduction of the retailer’s
optimal second-period decision into the manufacturer’s problem. Third, the optimal strategies
of Stages 3 and 4 are incorporated into the retailer’s first-period problem in Stage 2 to derive
7
the retailer’s optimal strategies of the period (p1, a1). Lastly, the manufacturer’s first-period
decisions (w1, s1) are obtained knowing the strategies of the previous stages.
We show in Appendix A that the conditions ensuring the strict concavity of the retailer’s
and manufacturer’s profit functions with respect to their second-period decision variables are
satisfied. However, the manufacturer’s and retailer’s profit functions with respect to their first-
period decision variables could be concave or not. As a result, an interior equilibrium that
maximizes the channel members’ objective functions may or may not be attained, depending
on the values of some model parameters. Further, we characterize the corner solutions and
list the conditions that ensure that the problem admits an interior solution (Scenario I). The
corner solutions correspond to s1 = 0, a situation where the manufacturer does not provide
any cooperative advertising support to the retailer in the first period (Scenario II), and a1 = 0,
a situation where the retailer does not undertake any advertising (Scenario III). Our analysis
focuses on these two corner solutions together with the interior solution, which are relevant to
the planning of cooperative advertising programs over time as addressed in this paper. The
following proposition characterizes these three possible Stackelberg equilibrium solutions.
Proposition 1. There exist three Stackelberg equilibria that correspond to the following threescenarios:
1. Scenario I: Under conditions (B.1)-(B.10) in Appendix B, the retailer advertises andthe manufacturer offers an advertising support to the retailer in each period.
2. Scenario II: Under conditions (C.1)-(C.3) in Appendix C, the retailer advertises in eachperiod, but the manufacturer only supports the retailer’s second-period advertising.
3. Scenario III: Regardless of whether or not the manufacturer offers a cooperative adver-tising program in the first period, the retailer only advertises in the second period andreceives advertising support from the manufacturer.
Proof. See Appendices A, B and C.
The strategies in Appendix A show that, in Scenarios I and II, except for the second-period
cooperative advertising support rate, all other decision variables depend on the parameters α
and β, α ∈ (0, 1) and β ∈ (−1, 1). Therefore, some conditions have to be imposed on these
parameters to ensure positive strategies. These conditions are displayed in Figure 1 below.
Scenario I (Figure 1) is not feasible when retailer advertising produces relatively large
negative carryover effects, regardless of its short-term effects. Because retailer advertising
that produces large negative carryover effects significantly reduces the second-period baseline
demand, it is neither in the interest of the retailer to continuously undertake this type of
advertising nor in the interest of the manufacturer to support it (Jørgensen et al., 2003).
8
Also, the case where both the short-term and long-term effects of retailer advertising simul-
taneously take very large values (top right of Figure 1) is not feasible as well. Due to very
large advertising effects, advertising expenditures can reach levels that are not economically
sustainable to channel members. Between these two extremes, the retailer can continuously
invest in advertising in the two periods and the manufacturer can find it optimal to support
retailer advertising providing that it does not dramatically hurts long-term sales. As a matter
of fact, small damages to the brand image can be compensated by short-term incentives that
boost current sales. It is therefore not surprising to see that major automakers have coop-
erative advertising policies in place that prevent them from supporting promotional activities
that can substantially harm their brand. For instance, Honda Canada Inc. prohibits retailer
advertising that implies a distressed sales environment, and requires retailer advertising to
have a minimum of two benefit descriptions for the featured products, and to properly use its
logo.
α
β
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1−1
−0.8
−0.6
−0.4
−0.2
0
0.2
0.4
0.6
0.8
1
Both Scen. I and II are feasible
Only Scen. IIis feasible
Both Scen. I andII are unfeasible
Only Scen. IIis feasible
Both Scen. I andII are unfeasible
Figure 1: Feasible and unfeasible regions for scenarios I and II.
9
Scenario II (Figure 1) is not feasible mainly when retailer advertising combines small
short-term effects with negative carryover effects. Current advertising-induced sales are not
enough to counterbalance the decrease of the second-period baseline demand. As above,
extreme cases of large short-term and positive long-term effects are also unfeasible because
of cost implications. Observe however that retailer advertising that highly impacts short-
term sales can be undertaken even if it has large negative carryover effects. This is because
the manufacturer’s advertising support to the second-period retailer advertising contributes
to increasing retailer advertising and consequently helps to mitigate the negative carryover
effects of the first-period advertising. The second-period cooperative advertising program
will become less attractive if the model is expanded to three periods and the second-period
advertising negatively affects the third-period sales.
In Scenario III, the two players’ strategies are all positive regardless of the parameter
values, which means that players can play this equilibrium with any type of retailer advertising.
Particularly, the first-period retailer’s and manufacturer’s prices are positive as expected in
a simple pricing game. The second-period retailer’s and manufacturer’s prices as well as the
second-period retailer’s advertising all depend exclusively on the short-term effects of retailer
advertising. The retailer’s decision not to advertise in the first period renders irrelevant the
carryover effects of advertising in this scenario.
Summarizing, the manufacturer offers a cooperative advertising program in the first and
second periods only under some specific conditions related to the effects of retailer advertising.
Any corner solution equilibrium chosen does not lead to an effective cooperative advertising
program in the first period. This is because either the manufacturer endogenously finds it
not optimal to support the retailer’s advertising in the first period mainly due to its negative
carryover effects (Scenario II) or the retailer does not undertake advertising to reduce her
advertising expenses over the two periods (as in Scenario III). The fact that the retailer
does not undertake local advertising nullifies any cooperative advertising program that the
manufacturer may consider implementing. Hereafter, we will therefore assume that sIII1 = 0.
4. Players’ strategies over time
We now investigate whether or not the manufacturer’s and retailer’s decisions change from
the first to the second period. The findings of the comparisons of the players’ strategies for
the two periods under each of the three scenarios are summarized in Proposition 2.
Proposition 2. The players’ decisions in the first and second periods compare as follows:
1. In Scenario I: For any α ∈ (0, 1),
10
(a) If β = 0, then sI1 = sI2 = 1/3, aI1 = aI2, wI1 = wI
2, pI1 = pI2.
(b) If β ∈ (0, 1), then sI1 > sI2 = 1/3, aI1 > aI2, wI1 < wI
2, pI1 < pI2.
(c) If β ∈ (−1, 0), then sI1 < sI2 = 1/3, aI1 < aI2, and wI1 and pI1 can be greater or
lower than wI2 and pI2, respectively.
2. In Scenario II: For any α ∈ (0, 1),
(a) If β = 0, then wII1 < wII
2 , pII1 < pII2 , aII1 < aII2 .
(b) If β ∈ (0, 1), then wII1 < wII
2 , pII1 < pII2 and aII1 can be greater or lower than aII2 .
(c) If β ∈ (−1, 0), then pII1 < pII2 and wII1 and aII1 can be greater or lower than wII
2
and aII2 , respectively.
(d) sII1 = 0 < sII2 = 1/3.
3. In Scenario III: For any α ∈ (0, 1), β ∈ (−1, 1), the following inequalities apply:
aIII1 = 0 < aIII2 , sIII1 = 0 < sIII2 = 1/3, pIII1 < pIII2 , wIII1 < wIII
2 .
Proof. The results of the comparison for the first and second scenarios can be easily derived
using Mathematica 10.0.
The findings of Scenario I critically depend on the short-term and long-term effects of
retailer advertising as illustrated in Figure 2 (In this figure and hereafter, UF denotes the
unfeasible region of the corresponding scenario). In this scenario, when the first-period retailer
advertising does not carry over to the second period, all the players’ decisions of the two periods
are identical. This result is intuitive. Consumer demands in the two periods are independent
when β = 0. The two-period game in this particular case can be considered as two successive
static games in a stable environment. Players’ strategies remain, therefore, unchanged over
time. Such strategies are comparable to steady-state strategies in differential games-based
models (e.g., Jørgensen et al., 2000; Sigue and Chintagunta, 2009).
On the other hand, unlike steady-state strategies, when retailer advertising positively im-
pacts on the second-period demand (β > 0), the retailer invests more in advertising and the
manufacturer supports a bigger share of retailer advertising in the first period to expand the
second-period baseline demand.The expanded second-period demand allows both the manu-
facturer and retailer to respectively increase their second-period wholesale and retail prices.
When retailer advertising harms long-term sales (β < 0), as expected, the retailer invests less
in advertising and the manufacturer provides a smaller cooperative advertising support rate
in the first period to limit adverse effects of retailer advertising. Less intuitive, however, is the
finding that the two players may charge higher or lower wholesale and retail prices in either
period of the game depending on the magnitude of both the short-term and negative long-term
11
α
β
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1−1
−0.8
−0.6
−0.4
−0.2
0
0.2
0.4
0.6
0.8
1
p1I <p
2I , w
1I <w
2I
p1I <p
2I , w
1I <w
2I
UF
p1I >p
2I , w
1I >w
2Ip
1I <p
2I , w
1I >w
2I
UF
Figure 2: Scenario I: Comparison retail and wholesale prices for the first and second periods.
effects of retailer advertising (Figure 2). Because the baseline of the second-period demand is
smaller, the two players charge higher prices in the first period as long as the damage on long-
term sales is relatively small. Otherwise, players need to set lower first-period prices to sell
as much as possible in the first period to mitigate pronounced negative long-term advertising
effects in the second period, especially when consumers are very sensitive to current effects of
retailer advertising.
The findings of Scenario II, in which the manufacturer does not find it optimal to support
the first-period retailer advertising, are illustrated in Figure 3.
All players’ decisions take higher values in the second period when advertising does not im-
pact on the second-period demand (β = 0). The change in the players’ strategies in the second
period is explained by the offering of a cooperative advertising program by the manufacturer,
which helps to increase retailer advertising and stimulate consumer demand. This finding is
consistent with the current static games literature in bilateral monopoly, which supports the
12
α
β
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1−1
−0.8
−0.6
−0.4
−0.2
0
0.2
0.4
0.6
0.8
1
w1II<w
2II, a
1II>a
2II
w1II<w
2II, a
1II<a
2II
w1II>w
2II, a
1II<a
2II
UF
UF
Figure 3: Scenario II: Comparison wholesale prices and advertising investments for the first and second periods.
view that cooperative advertising increases retailer advertising and consumer demand (Aust
and Buscher, 2014; Jørgensen and Zaccour, 2014). Again, because the decisions and demand
functions in the two periods are independent, this case corresponds to two successive static
games, where the manufacturer unilaterally changes his set of decision variables from one
game to another.
A dynamic game is played when the first-period advertising produces long-term effects
(β 6= 0). As expected, when retailer advertising expands the second-period demand (β > 0),
the two channel members charge higher second-period prices. Surprisingly, the retailer may
invest more in advertising in the first period than in the second period, especially when
advertising heavily stimulates long-term sales and its effects on current sales are relatively
small. Although the manufacturer does not offer any cooperative advertising support in the
first period in this scenario, the relative effectiveness of advertising over time can still motivate
the retailer to invest more in advertising in the first period than in the second period where
13
cooperative advertising is provided.
On the other hand, when retailer advertising negatively impacts on long-term sales (β < 0),
all second-period decisions are still higher than the first-period decisions, except in this case
that the manufacturer may reduce his second-period wholesale price, especially when the
negative long-term effect and the current effect of retailer advertising are high. The reduction
of the second-period wholesale price can be considered as an additional incentive, on top of
the cooperative advertising support, given to the retailer to support additional marketing
activities in the second period. Such an extra incentive limits the damaging effect of the
first-period advertising as the retailer is given larger margins to be used at her own discretion.
The findings of Scenario III, in which the retailer does not undertake advertising in the first
period, reveal that regardless of the values of the parameters, the second-period decisions for
the two players take larger values. The demand functions in the two periods are independent as
the retailer does not advertise in the first period. This scenario corresponds to two successive
static games, where the retailer unilaterally changes her set of decision variables from one game
to another. The second-period retailer’s advertising combined with the cooperative advertising
support provided by the manufacturer help to expand consumer demand during this period.
As a result, both the manufacturer and retailer respectively increase their wholesale and retail
prices.
5. Selecting an equilibrium
The goal of this section is to endogenously evaluate, from each player’s perspective, which
of these three equilibria provide the best profits. The comparisons of the manufacturer’s and
retailer’s optimal profits for the three different equilibria are presented in Figure 4. Analytical
expressions are huge and cannot be presented here (these expressions are presented in Ap-
pendix A). The region where the manufacturer and retailer obtain the greatest profits when
equilibrium i is applied (i = I, II, III) are respectively denoted by M i and Ri in Figure 4.
Focusing only on M i, Figure 4 shows that, depending on the area in the parameter space,
the manufacturer can select either one of the three scenarios. For instance, the manufacturer
prefers Scenario III because it is either the only feasible scenario in certain areas or it secures
the best profits when, at least, another scenario is feasible. In the latter case, Scenario III
is the manufacturer’s best choice when retailer advertising either harms long-term sales and
is not very effective in stimulating short-term sales, or is not very effective in boosting both
short-term and long-term sales.
To understand the manufacturer’s preference, consider for example, the case where retailer
14
advertising does not affect long-term sales (β = 0), as can be seen in Appendix D, the two
players’ second-period decisions are identical in the three scenarios. The difference among
these scenarios stems from the decisions of the first period and the resulting demands. The
retailer advertises in the first period in Scenarios I and II and these investments can lead
to higher or lower first-period demands in these two scenarios compared to Scenario III.
Particularly, the manufacturer makes more profits in Scenario III than in Scenarios I and II,
when the addition of retailer advertising to the mere pricing game played in the first-period
of Scenario III reduces consumer demand. This happens in situations where the short-term
effects of retailer advertising are not large enough to compensate its induced price increases.
Otherwise, when the short-term effects of retailer advertising are very large, the manufacturer
prefers Scenario I, which generates the highest demand.
Scenario III becomes even more interesting than the other two, when the first-period
advertising harms long-term sales (β < 0). Because the retailer does not invest in advertising
in the first period in Scenario III, the baseline demand in the second period is not negatively
affected and this gives the possibility to all channel members to charge higher prices and earn
more profits in the second period. The opposite applies to the situation where the first-period
advertising expands the second-period baseline demand (β > 0) as Scenarios I and II become
more attractive in the second period. The difference between these two scenarios is explained
by the manufacturer’s first-period cooperative advertising in Scenario I and its impact on both
pricing and advertising decisions in the two periods.
Considering exclusively Ri in Figure 4, it is easy to see that the retailer prefers Scenario III,
either at the bottom of the figure when advertising significantly damages the second-period
sales or the adoption of a continuous advertising schedule becomes extremely expensive due
to large short-term and long-term effects (top right of Figure 4). Observe that in these
two areas, Scenario I is not feasible. Between these two extreme regions, where the three
scenarios are feasible, the retailer prefers Scenario I in which she advertises and benefits from
the manufacturer’s support in each period. Thus, unless the retailer finds it optimal not to
advertise in the first period (Scenario III), she is better off when the manufacturer shares
her advertising expenses than when he does not. This is because the first-period cooperative
advertising program stimulates retailer advertising and may boost both the first- and second-
period sales, especially when advertising positively affects long-term sales.
Looking at M i and Ri together in Figure 4, one can identify the areas where the two
channel members have similar/dissimilar preferences.
Both the manufacturer and retailer agree to implement Scenario III (regions M III , RIII)
either at the extreme bottom of Figure 4 when advertising significantly damages the second-
period sales or advertising has simultaneously large short-term and long-term effects at the
top right of Figure 4. The two channel members may also agree to implement Scenario I
15
α
β
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1−1
−0.8
−0.6
−0.4
−0.2
0
0.2
0.4
0.6
0.8
1
MIII,RIII
MIII,RI
MI,RI
MI,RIII
MII,RIII
MIII,RIII
Figure 4: Channel partners’ agreement regions.
(region M I , RI), mostly in areas where the long-term effects of advertising are positive. There
are three areas of disagreement: the manufacturer and retailer respectively prefer Scenario
III and Scenario I (M III , RI), the manufacturer and retailer respectively prefer Scenario II
and Scenario III (M II , RIII), and the manufacturer and retailer respectively prefer Scenario
I and Scenario III (M I , RIII). These situations may create conflicts within a channel unless
specific actions are taken to align the interests of the parties. As the channel leader, the
manufacturer may also specify up front which type of advertising qualifies for his cooperative
advertising program or supply recommended layouts and copies, forcing the retailer to invest
in the type of local advertising that best meets the interests of the two parties. For instance,
Toyota prohibits dealer advertising that features vehicle prices below invoice that generally
has a strong short-term effect, but harms sales in the long run. Mitsubishi Motors offers to
its dealers an online advertising planner, which allows them to create their own ads without
having to hire an advertising agency and to use pre-prepared layouts and creative copies
16
(Jackson, 2004).
6. Conclusion and discussion
This paper analyzes how the manufacturer and retailer, in a bilateral monopoly, should
respectively set their cooperative advertising support rates and levels of investment in local
advertising over a two-period planning horizon. Three advertising arrangements are endoge-
nously identified: The manufacturer offers cooperative advertising support and the retailer
advertises in each period (Scenario I); the retailer undertakes advertising in the first and sec-
ond periods, while the manufacturer only supports the second-period advertising (Scenario
II); and the retailer only advertises in the second period and receives support from the man-
ufacturer (Scenario III). The decision to implement either one of the three scenarios critically
depends on the short-term and long-term effects of retailer advertising. The two channel
members can agree to play either Scenario I or III. The analysis of the players’ strategies over
time for each of the three scenarios reveals that the long-term effects of retailer advertising
influence how the players set their decisions from one period to another in Scenarios I and II.
Conversely, the long-term effects of advertising do not affect the players’ strategies in Scenario
III as the retailer does not advertise in the first period.
The findings of this research extend the existing literature and offer a integrative framework
that better explains observed practices in the business world and complement the findings
of previous static and differential games-based models. Static game models overlook the
long-term effects of advertising. Our findings suggest that the long-term effects of retailer
advertising should not be ignored, as they affect the players’ strategies and their changes over
time. On the other hand, differential games-based models mainly prescribe continuous retailer
advertising and constant cooperative advertising support rates over time (e.g., Jørgensen et
al., 2003; He et al. 2011; Sigue and Chintagunta, 2009; Zhang et al., 2013). We found
that constant cooperative advertising rates over time are not optimal when the first-period
advertising impacts on the second-period demand (Scenario I).
Another distinctive contribution of this research is that players can endogenously change
their decisions to invest or not to invest in advertising over time to maximize their respective
profits over the two periods. Channel members should therefore examine alternate advertising
schedules such as pulsing, as described in the advertising literature (Dube et al., 2005; Mahajan
and Muller, 1986; Mesak and Ellis, 2009; Sasieni, 1989; Villas-Boas, 1993). As a matter of
fact, the existence of the two corner equilibria shows that, the manufacturer can postpone
his cooperative advertising program to a later start, while the retailer pursues a continuous
17
advertising strategy from the beginning (Scenario II). Also, the retailer can find it optimal to
delay advertising for the second period regardless of whether or not the manufacturer offers
to support first-period advertising (Scenario III). While there are many reasons for which
retailers may not participate in cooperative advertising programs (a few examples are: lack of
money to afford one’s share of expenses, lack of information, and exogenous factors associated
with a cooperative advertising program), this research claims that the eligibility requirements
that define the type of advertising supported by manufacturers such as Honda, Mitsubishi,
and Toyota in the automobile industry should not be overlooked. Both the short-term and
long-term effects of advertising are critical in determining whether or not retailers should
prefer Scenario III.
The findings of this research also have practical relevance and can help managers to bet-
ter design their pricing and advertising decisions over time. The conventional wisdom in the
channel literature is that retailer advertising mainly stimulates short-term sales. Previous
differential games-based models have acknowledged that retailer advertising may have various
long-term effects (Jørgensen et al., 2000, 2003). Empirical research in the automobile industry
also supports the existence of retailer advertising carryover effects (Herrington and Dempsey,
2005). We apply this knowledge base to a two-period game setup, which can easily be visu-
alized in a business context and help to deal more effectively with time-limited cooperative
advertising contracts. Our findings support the view that when cooperative advertising pro-
grams are offered one after the other as observed in many industries, their optimal support
rates can change from one period to another depending on the type of retailer advertising
that the manufacturer wants to stimulate (Kachadourian, 2005). In addition, manufacturers
can design their cooperative advertising programs to find the right balance between stimulat-
ing short-term sales and maintaining the value of their brands. Similarly, manufacturer can
gradually decrease their support to brand-enhancing advertising activities for new products
as they move towards maturity.
We have simplified our model specification to derive meaningful analytical results. Some
of our assumptions can be relaxed to deal with more complex situations. For instance, the
manufacturer may be given the possibility of also undertaking national advertising to com-
plement the retailer’s advertising efforts. Consumer sensitivity to price and advertising may
change from one period to another as it is often the case for seasonal products (Martın-
Herran et al., 2010). An expansion to a three-period game could be considered to investigate
the long-term effects of the second-period retailer advertising and how they impact on both
the players’ strategies and profits. Competition could be added to take into account pricing
18
and advertising horizontal interactions between manufacturers and/or retailers.
References:
[1] Aust, G., & Buscher, U. (2014). Cooperative advertising models in supply chain manage-
ment: A review. European Journal of Operational Research, 234, 1–14.
[2] Berger, P. D. (1972). Vertical cooperative advertising ventures. Journal of Marketing
Research, 9, 309–312.
[3] Chintagunta, P., & Jain, D. C. (1992). A dynamic model of channel member strategies
for marketing expenditures. Marketing Science, 11(2), 168–188.
[4] Chu, W., & Desai, P. S. (1995). Channel coordination mechanisms for customer satisfac-
tion. Marketing Science, 14, 343–359.
[5] Cole, A. (2015). Toyota asking dealers to stop advertising below invoice, world not over
yet. The Truth About Care:http://www.thetruthaboutcars.com/2015/08/toyota-asking-
dealers-stop-advertising-invoice-world-not-yet/ Accessed on June 10, 2016.
[6] Dube, J-. P., Hitsch, G. J., & Manchanda, P. (2005). An empirical model of advertising
dynamics. Quantitative Marketing and Economics, 3, 107–144.
[7] He, X., Krishnamoorthy, A., Prasad, A., & Sethi, S.P. (2011). Co-0p advertising in
dynamic retail oligopolies. Decision Sciences, 42(4), 923-944.
[8] Herrington, J. D., & Dempsey, W. A. (2005). Comparing the current effects and carryover
of national-, regional-, and local-sponsor advertising. Journal of Advertising Research,
March, 60–72.
[9] Honda Canada Inc. (2010). Co-op advertising. http://motorcycle.honda.ca/Content/mcpead-
builder.ca/3660c2dd-cfb9-45d4-bd0b-a489d15e2697/AdBuilderAsset/Honda Co Op Policy
EN v2.pdf). Accessed on July 16, 2016.
[10] Huang, Z., & Li, S. X. (2001). Co-op advertising models of manufacturer-retailer supply
chains: A game theoretic approach. European Journal of Operational Research, 135, 527–
544.
[11] Jackson, K. (2004). Mitsubishi shakes things up. Automotive News, April 5.
19
[12] Jørgensen, S., Sigue, S. P., & Zaccour, G. (2000). Dynamic cooperative advertising in a
marketing channel. Journal of Retailing, 76, 71–92.
[13] Jørgensen, S., Sigue, S. P., & Zaccour, G. (2001). Stackelberg leadership in a marketing
channel. International Game Theory Review, 3(1), 1–14.
[14] Jørgensen, S., Taboubi, S., & Zaccour, G. (2003). Retail promotions with negative brand
image effects: Is cooperation possible? European Journal of Operational Research, 150,
395–405.
[15] Jørgensen, S., & Zaccour, G. (2014). A survey of game-theoretic models of cooperative
advertising. European Journal of Operational Research, 237, 1–14.
[16] Kachadourian, G. (2005). Wild dealer promotions pull in potential buyers. Automotive
News, 79(6132), 34-36.
[17] Karray, S. (2013). Periodicity of pricing and marketing efforts in a distribution channel.
European Journal of Operational Research, 228, 635–647.
[18] Karray, S. (2015). Cooperative promotions in the distribution channel. Omega, 51, 49-58.
[19] Karray, S., & Amin, S.H. (2015). Cooperative advertising in a supply chain with retail
competition. International Journal of Production Research, 53 (1), 2015.
[20] Karray, S., & Zaccour, G. (2006). Could co-op advertising be a manufacturer’s counter-
strategy to store brands? Journal of Business Research, 59, 1008–1015.
[21] Li, S., Huang, Z., Zhu, J., & Chau, P. (2002). Cooperative advertising game theory and
manufacturer-retailer supply chains. Omega, 30(5), 347-357.
[22] Lilien, G. L., Kotler, P., & Moorthy, K. S. (1992). Marketing Models. New Jersey: Prentice
Hall.
[23] Mahajan, V., & Muller, E. (1986). Advertising pulsing policies for generating awareness
for new products. Marketing Science, 5(2), 89–106.
[24] Martın-Herran, G., & Sigue, S. P. (2011). Prices, promotions, and channel profitability:
Was the conventional wisdom mistaken? European Journal of Operational Research,
211(2), 415-425.
20
[25] Martın-Herran, G., Sigue, S. P., & Zaccour, G. (2010). The dilemma of pull and push-
price promotions. Journal of Retailing, 86(1), 51–68.
[26] Mesak, H. I., & Ellis, T. S. (2009). On the superiority of pulsing under a concave advertis-
ing market potential function. European Journal of Operational Research, 194, 608–627.
[27] Mitsubishi Motors. (2012). Individual co-op assitance pro-
gram: http://tempe.acbcoop.com/paranet60/clientdocs/mitsubishina-
multipoint/2012coopguidelines04-06.pdf/ Accessed on July 21, 2016.
[28] Sasieni. M. W. (1989). Optimal advertising strategies. Marketing Science, 8(4), 358–370.
[29] Sigue, S. P. (2008). Consumer and retailer promotions: Who is better off? Journal of
Retailing, 84(4), 449–460.
[30] Sigue, S. P., & Chintagunta, P. (2009). Advertising strategies in a franchise system.
European Journal of Operational Research, 198 (2), 655–665.
[31] Szmerekovsky, J.G., & Zhang, J. (2009). Pricing and two-tier advertising with one man-
ufacturer and one retailer. European Journal of Operational Research, 192, 904–917.
[32] Villas-Boas, J. M. (1993). Predicting advertising pulsing policies in an oligopoly: A model
and empirical test. Marketing Science, 12(1), 88–102.
[33] Wolfram Research, Inc. (2010). Mathematica Edition: Version 10.0. Wolfram Research,
Inc. Champaign, Illinois.
[34] Xie, J., & Ai, S. (2006). A note on “cooperative advertising, game theory and
manufacturer-retailer syupply chains”. Omega, 34, 501–504.
[35] Yan, R. (2010). Cooperative advertising, pricing and firm performance in the e-marketing
age. Journal of the Academy of Marketing Science, 38, 510–519.
[36] Yan, R., Cao, Z., & Pei, Z. (2016). Manufacturer’s cooperative advertising, demand
uncertainty, and information sharing. Journal of Business Research, 69, 709-717.
[37] Yan, R., & Pei, Z. (2015). The strategic value of cooperative advertising in the dual-
channel competition. International Journal of Electronic Commerce, 19(3), 118-143.
[38] Zhang, J., Gou, Q., Liang, L., & Huang, Z. (2013). Supply chain coordination through
cooperative advertising with reference price effect. Omega, 41, 345–353.
21
[39] Zhang, J., Cou, Q., Li, S., & Huang, Z. (2015). Cooperative advertising with accrual
rate in a dynamic supply chain. Dynamic Games and Applications. Published online 04
August 2015. DOI: 10.1007/s13235-015-0165-z
22
top related